Dan GillmorAug 9, 2023

Authors going berserk over machines **reading** their work are completely missing the point. There is nothing wrong with reading.

There are some legitimate worries about what machines do after they read -- and store -- the material.

But as @mmasnick points out in this latest case of misguided author freakout, that's a nuance that creators -- and journalists -- can't be bothered to consider:

https://www.techdirt.com/2023/08/08/the-fear-of-ai-just-killed-a-very-useful-tool/

The Fear Of AI Just Killed A Very Useful Tool

I do understand why so many people, especially creative folks, are worried about AI and how it’s used. The future is quite unknown, and things are changing very rapidly, at a pace that can feel out…

Techdirt
Show threadJude
@dangillmor @mmasnick The fear is not so much about the LLM technology itself but the companies that empower it. There have been years and years of watching tech monopolies form and engage in abusive anti-competitive behavior, driving out many of benefits to users and descending into the now familiar degradation. #enshittification
Aug 9, 2023 at 1:08amWeb
Show thread$8TrollAug 9, 2023

@sgraffito @dangillmor @mmasnick

In order to train AI a copy of the text must be stored. This copy might not comply with copyright or fair use doctrine.

But it isn't just reading, it is a form of copying the content (represented by probabilities in the AI) that will lead the output to mimic inputs in a way that is not covered by parody or satire.